Oooh, shiny thing!

While they have plenty of limitations, tag clouds are excellent for outlining the relative importance of many concepts at once. They're also visually very attractive, a lot more eye-catching than a uniform paragraph of text or a bullet-point list.

I'm having trouble coming up with any arguments on what a 3D tag cloud is actually good for (if pushed I'd mumble something about using the extra dimension to show time information) but boy, is it it pretty! Awesome job by Roy Tanck, and thanks to Logic Nest for finding it.

Step 1- Network Analysis, Step 2- Profit


My friend Mrinal just pointed me towards someone I should have found long ago. Rob Cross is a management professor at the University of Virginia who's applied social network analysis to improve all sorts of companies. The part I like the most is that his case studies are strongly focused on concrete results and money saved. Here's some of the things that he's used network analysis for:

– Cloning a successful team by duplicating its structure with another set of people
– Fostering innovation by identifying brokers who spread improvements across the organization
– Spreading best practices by integrating groups within the company working on similar problems
– Building connections between sales teams so they can boost revenue by cross-selling

I've ordered his two books, Driving Results Through Social Networks and Hidden Power of Social Networks:

If you're looking for a layman's guide to what he's able to do for companies, his newspaper articles are a good place to start.

All of his analysis, in common with other practitioners like Ronald Burt, is based on self-reporting and surveys. I'm pretty convinced that many of these same techniques will be even more interesting and detailed if you apply them to implicitly generated communications data. Your email inbox is going to be a more accurate and comprehensive record of who you communicate than your recollection.

A better source of Enron’s emails in PSTs


Photo by Irish Typepad

Dr John Wang [Update- sorry, wrong John Wang!] has just started a new site called, dedicated to developing and refining the Enron email dataset. It’s off to a cracking start, offering all the Enron emails as 148 PST files, one for each ‘custodian’ (informally each mail user). I did my own PST conversion, but it was primarily so I had a large data set to load onto an Exchange server and test Mailana against. John’s version is much closer to the original source data, and so will be more of a real-world test for applications.

I’m really pleased John has put this together, it will be a boon to anyone looking at doing heavy-duty email data-mining. I can’t wait to see what else the project produces.

Who are the email innovators?


Photo by Cayusa

Founder of ClearContext, Deva's been pushing the boundaries of email productivity for years with his Outlook plugin. There's too many clever ideas in there to list, and who can argue with a guy who got his seed-funding playing online poker?

T.A.'s behind Gist. Along with Steve Newman, they're building a startup focused on linking external information to email contacts, breaking down the silo that surrounds email, a cause dear to my own heart. I'm jealous of the name too, very appropriate and memorable.

Former Engineering VP of Xobni, Gabor's now beavering away on a new email startup. Not many details have emerged yet, but I know it will be worth watching.

Sherman Dickman

Co-founder of Postbox with Scott MacGregor, Sherman's building an entirely new email client, with a radically new interface for working with all the information we've accumulated in our inboxes. The dominant UI we have at the moment for mail messages is essentially a DOS directory listing, Postbox (and Xoopit with some similar ideas) gives you a truly graphical user interface for mail.

Joshua is the founder of OtherInbox, another company trying to move email away from being a flat list. Rather than reimagining the core interface, they're using multiple email addresses and automated-email recognition to present all the information that Facebook and other services send via email in a much simpler form. And by removing that bacn from your messages, they return email to its roots as a person-to-person medium, one of the big reasons that Facebook mail is so pleasant to use.

CEO of Inboxer, Roger's one of the few people I know innovating in email on the server side. Since administrators make the purchase decisions there, not end users, a lot of the product is driven by regulatory and security concerns, but they also manage to offer some interesting options for searching and recovering attachments.


Who am I missing? Let me know…

Runners are crazy


Photo by Giorgio Zanetti

My friend Howard Cohen, one of the few people to complete the 70 mile Backbone Trail in a day, is off to North Africa next month to take part in the Libyan Challenge. Running 120 miles across the desert, unsupported, he's part of the first American team ever to enter the race.

I like to think I'm fairly fit, but Howard and the other ultra-runners are in a class of their own. It's awe-inducing to see how far they can stretch a human body, and the mental toughness they need to get through their ordeals is amazing. It takes some inspired insanity to tackle challenges like these.

Stop by Howard's site and wish him the best, and keep an eye on the offical website if you want to follow their progress starting February 24th.

Is Boulder a community or a clique?


The comments on Micah's post about the Boulder tech scene got me thinking about the difference between a community and a clique. On one level, a clique is any community you're not in. All social groups offer advantages to their members and have barriers to entry, which can make them infuriating to outsiders.

Communities work because they distribute the cost of evaluating strangers across all the members, and magnify the cost of offending any member by spreading information about transgressions. In other words, social networks generate reputations. Reputations aren't perfect, but they're harder to fake than any other signal we've got about trustworthiness.

So, we need groups, but we all know from high school how they can be used for evil. What distinguishes a good community from a clique?

Meritocratic. Inclusion and status should be determined by something more or less objective.

Open. There should be established, well-known ways to become part of the group.

Positive. The community should be defined by who they include, not who's excluded.

In my experience, Boulder's tech scene is all of these. Sure, there's well-connected people, and others who are less in the loop, but the central people tend to be there because they're heavily involved in events, or have a long track record in startups. There's a lot of events and programs like TechStars,, new tech meetups, hackspace and Ignite Boulder that are good entrypoints to the community.

A shocking expose of the venture funding process!

If you want to understand how VCs really decide which startups to fund, or how entrepreneurs arrive at their sales forecasts, watch this stunning series from the San Diego Venture Group. If nothing else, you’ll learn the best way to approach VCs (hint- it involves assembling a massive spreadsheet of hundreds of email addresses and sending nightly emails to them all until they agree to give you money).

Thanks to Startup Lawyer and Brandon Zeuner for alerting me to this fine example of investigative journalism.

How to view the MAPI/RPC documentation online

No connection to the post for once, I just can't resist an angry lemur

Photo by Law Keven

Microsoft recently released the documentation for the secret protocol Outlook uses to communicate with Exchange. Yeay! Unfortunately they released it as a large number of PDFs in a zip file. Boo!

I've been using them for my work on Mailana, but having to use local file searching or manual browsing through all these documents rather than my usual web search has slowed me down. Today I finally bit the bullet, ran them through a PDF batch converter to get HTML, and put them online at

In a few days all that lovely information should show up in Google searches, and the site search should work too. Thanks to Darren Hoyt for the simple site-search HTML and PDF Bean for the conversion tools.

It's a shame that Microsoft's own documentation is so unfriendly to the web. They often change links without implementing forwarding, so often old blog or forum posts lead nowhere, and some documentation like this is only available as unsearchable downloads. Of course Apple can be worse, requiring logins before you can get at a lot of the resources, and the mailing list search tools are worthy of a geocities page ten years ago. It's funny how lost I feel when I'm researching an area that's invisible to Google, it really has become half of my brain.

What makes a great salesman?


Photo by The Life of Bryan

One of the things I really suck at is selling. Part of it is growing up in Britain with the belief it's about tricking people into buying things they don't need. Being a professional engineer didn't help either. Most programmers are simply baffled that customers don't simply get that their product is better. Why do they need some highly-paid guy in a suit to get involved?

As I got older, I realized that every job is a sales job. To get anything done, you need to persuade a whole bunch of internal and external people to help. Now I'm running a startup, and that's all about selling the idea to everyone I need to deal with; investors, business partners, employees and customers.

I've looked around for role models. My favorite so far is an infomercial host called Ron Popeil. I can imagine my British friends cringing because he's almost a caricature, but this profile by Malcolm Gladwell opened my mind to both how much dedication he has, and how effective he's been. So, what are his secrets?

Feedback and measurement

He's from a family with a long tradition of selling on street corners. At the end of an afternoon, they'd know exactly how much they'd brought in. That gave them a guide they could use to figure out what worked and what didn't. The infomercials followed the same principles, with real-time graphs showing how many people were calling and ordering.

This might sound obvious, but as Dim Bulb repeatedly demonstrates, most TV advertising is driven by the intangible idea of brand, with no idea what's actually working or failing. It's like the difference between the Greek philosophers building elaborate theories on how the universe works, and experimental science that's able to test ideas.

I've become a fanatic on trying to measure everything I can about my communications, keeping track of who I've talked to and when, measuring which pages and posts get visitors, using online ad experiments to gather survey data. Without that foundation, I'll never be able to improve.

Involvement in design

Ron actually built and designed his products in his kitchen. The goal was to create something that would sell itself. Too often, there's a distance between sales teams and the people who build the products. They might have some voice in the planning stages, but they're shut out during the implementation and expected to take whatever the result is and sell it.

Since I keep swapping hats between selling and building, you'd think I wouldn't have this problem. It's funny though, I often get caught up in the geeky coolness of the technology, and lose sight of what people are willing to pay for. The lesson I took from Ron's example here was to keep asking myself what problem every feature I'm working on is actually solving.

Product focus

In the infomercials, the camera quickly focuses on the gadget, and stays there. It's not about the personality of the salesman, it's all about what the device can do. There's an anecdote in Gladwell's story about a showdown between a few salesmen at a trade show. Frosty Wilson was charming and persuasive, everything you'd imagine a great salesman should be, but Ron and his partner both sold twice as much by making the product the star.

In my job, I've learnt it's best to cut to the demo as quickly as possible, and then let people try the prototype themselves. Nobody wants to sit and listen to a lecture, it's much more compelling to see what it can do rather than be told.

Fervent belief

Ron really, truly believed that the products he was selling would make his customers lives better. He sounds like Steve Jobs when he's hammering away at the smallest details of every design, making sure that everyone gets an experience they'll be delighted with. It's not just about getting their money, it's his purpose.

Luckily I am insanely convinced that what I'm building will change the way we work. I've found I'm most effective when I can just informally rant about all the amazing possibilities rather than sticking to a script.

Your real social network


Photo by Law Keven

Stowe Boyd covered an interesting paper on social networks and concluded "the apparent, superficial social network based on following and followers conceals a deeper, sparser social network". Every current service has an incredibly primitive representation of your relationships. You're either friends with somebody, or you're not. Here's what''s missing:

Strength. There's no way to specify how close you are to somebody else.
Time. Is the friendship long-lasting? Have you talked recently?
Context. What other friends is this friend close to? Which circles do they move in?

It's well known that you can use communication data to answer these questions. This implicit approach is better than trying to get people to enter this information manually because:

Convenience. Nobody wants to spend time doing data entry and house-keeping on their network. Doing it automatically solves that problem.
Reliability. You can objectively measure how many emails somebody has sent you, and how many you've returned to them. This removes the subjective element that creeps in if you're asked to rate the strength of a relationship on an arbitrary scale. It also removes the temptation to exaggerate your closeness to someone influential.

So why hasn't anyone done this? There's massive technical barriers to overcome before you can access large stores of email, and big privacy issues. I'm convinced they can be overcome, and that's what I'm doing with Mailana. If you want to see the sort of detailed social graph I'm talking about, Boulder Twits is using the same backend as my email analysis system.